Transcutaneous Implantation Methods for Improving the Long-Term Performance of Glucose Sensors in Rats

Translation of sensor design and function in animal models to human use is an ongoing challenge due to tissue anatomical and physiological differences between species, even at presumably analogous implant locations. Nevertheless, preclinical testing of sensors for long-term glucose monitoring in animals is required for evaluating sensor function in order to improve sensor design. Long-term glucose sensor testing in common laboratory animals (e.g., mice and rats) is especially difficult due to their small size, as well as limited site availability for sensor placement without disturbance or removal by the subject. However, improvements in sensor design and implantation methods to improve sensor survival in these animals could accelerate our understanding of the role of tissue reactions to sensor components, as well as allow reliable testing of biomaterials and various drug or growth factor delivery systems to potentially minimize or modulate tissue reactions. In this study, methods to secure a wire-type subcutaneous sensor in rats for a long period of time (ges28 d), utilizing new implantation techniques and devices were evaluated. Anchoring devices were incorporated into the sensor design and appropriate implantation methods were used to: (1) minimize potential membrane damage caused by animal motion; (2) prevent removal of the entire sensor or sensor wires by the animal; and (3) allow exterior access to wires for periodic sensor performance testing. The anchoring devices for securing sensors to the skin internally, which were sequentially investigated and improved (Protocol A to C), included a modified 22 gauge intravenous winged catheter (Protocol A), Silastic tubing (Protocol B) or silk suture loops held in place by Silastic tubing (Protocol C). The results show that after four weeks implantation, 60% (n = 10), 70% (n = 10), and 92% (n = 12) of the implanted devices survived (Protocols A, B, and C, respectively). Functional testing showed that 30% (n = 10), 40% (n = 10), and 58% (n = 12) of the sensors still worked well four weeks after implantation (Protocols A, B, and C, respectively). No infections were visibly evident at the sites of sensor implantation at any time during the testing period for all protocols. Protocol C shows promise as a viable method for future sensor studies because of the anchoring device's small size and because it was nearly impossible for rats to remove or damage the sensors.